Carpenter — Relationships between Classes of Arthropoda. 351 



Insects and Crustaceans — unless we arc willing to explain any like- 

 ness whatsoever as the result of " convergence." 



The concealed situation in which the Peripatids live might incline 

 us to the view that they have lost an originally firm exoskcleton. 

 Inscct-larvpe that live in wood are soft-shinned, while their allies in 

 the outer world are well armoui'cd. Yet there is an undegi'aded 

 aspect about a Peripatid that makes such a view hard to accept; and it 

 is more reasonable to regard the type as a very ancient one that 

 has come down, like certain Brachiopods, from a remote period, with 

 very little modification. But granting this, do the Peripatids really 

 help to bridge the gap between other Arthropods and Annelids ? The 

 soft skin, the simple eyes, and the segmental organs are really the only 

 distinctly Annelidan characters of the Malacopoda, and the force of the 

 last-named and most important of these is greatly weakened if Good- 

 rich's view ('97), accepted by Lankester ('00), be established, that the 

 coelomic ducts of Arthropods (to which category the segmental organs 

 of Peripatus must certainly be referred) have nothing to do with true 

 Annelidan nephridia. Certainly the legs of Peripatus resemble 

 Annelidan parapodia as little as they resemble Crustacean appendages. 

 On the whole, then, the Malacopoda are low-type Arthropoda, of 

 uncertain segmentation, but with the fundamental characters of the 

 phylum, and showing only a superficially Annelidan appearance. 



The question of the Annelidan ancestry of the Arthropods must 

 remain, then, a matter for speculation. To the present writer it 

 seems unsound morphology to compare closely the most highly- 

 developed class of worms (Chastopoda) with the most highly-organised 

 of all Invertebrates (Arthropoda). The presumption must always be, 

 in such cases, that each group has become specialised along its own 

 lines, and it is most unlikely that the one can have developed 

 directly from the other. Both may have diverged from a common 

 ancestiy; but the closed blood-vascular system and coelomic body- 

 cavity of the Chsetopoda point to the period of such an ancestry as 

 immensely remote. 



As contrasted with the vast difficulties involved in the transfonna- 

 tion of Polychaete worms into Phyllopods, the derivation of the 

 primeval Ai'thropods from iN'aupliif orm ancestors by a gradual increase 

 in the number of segments is perfectly simple ; and, before many years 

 have passed, zoologists are likely to revert to Miiller's theory of 

 Crustacean origins. The occurrence of the iN'auplius larva, or its 

 representative, in aU the great groups of Crustacea back to the Trilo- 

 bites raises the strongest presumption of some phylogenetic meaning ; 



R. I. A. PEOC, VOL. XXIV., SEC. K.] 2 F 



